Research On The Methods Of Imaging Quality Assessment Of Digital Remote-sensing Images

With the gradual development of photogrammetry and computer technology, digital remote -sensing image and image map have been applied to all areas more and more widely. Facing the massive image data of good or bad quality, data filtration is of great importance. For this reason people attach more attention to researches in the assessment of image quality. This dissertation has researched deeply both in theory and technique of the assessment of imaging quality of remote-sensing images. The main innovations are as follows:(1). Sum up the basic theory of digital remote sensing image quality assessment. Based on this, classical assessment algorithms such as information entropy, mean square method, SNR method are analyzed while pointing out their advantages and disadvantages;(2). Introduce a few conclusions of the human visual system to the assessment of image contrast, and design a new assessment method. The theory of image segmentation to the assessment of image contrast is introduced According to the characteristics of human vision that different regions attract different interests from human eyes a new method of image contrast assessment is designed. This new method also gives much consideration to the Surveying and Mapping field.(3). Clarity refers to the clear degree of the image's edge. In this dissertation and the designed assessment method of remote sensing image's clarity is around this concept. Traditional assessment methods in assessment of clarity incorporate all the pixels into the calculations without extracting the edge pixels of images. For this point, the designed assessment method of the clarity of images in this dissertation extracts the edge of the images firstly. Based on this, the subsequent applications of images are taken into the assessment of image clarity and the canny edge-detecting algorithm is introduced into the new clarity-assessing method instead of the other edge-detecting algorithms after considering the characteristics of the edge-detecting algorithms.(4). Analyze the relative theories and methods of general assessment of imaging quality and introduce the Image Quality Assessing System and its function modules.